Childhood obesity is a significant problem nationwide, but certain subgroups are at higher risk. This study targets one high risk subgroup, children living in a rural, low-income Appalachian area. This study will test the effectiveness of an innovative delivery model of health curriculum using trained high school-aged mentors compared to the usual condition of using adult group leaders. The content of the curriculum includes information regarding healthy eating and activity. Of primary interest is the delivery mode of the information, through a child and teen mentor relationship, compared to an adult group leader in a classroom setting. The curriculum used in this study combines concepts from Social Cognitive Theory, Theory of Planned Behavior, and Self Determination Theory. This between-group study compares a control (adult leader) and an experimental treatment (teen mentor) group. The target population is third and fourth grade students in a rural, low-income county in the Appalachian region. Children enrolled in an after-school program are eligible to participate. Tenth and 11th graders will participate as mentors. All participating elementary aged children will be randomly placed into either the control or experimental group. The difference between the pretest and posttest scores for independent variables: (a) nutritional knowledge, (b)attitudes towards eating healthy, (c) attitudes towards being physically active, (d) self efficacy for eating healthfully, (e) self efficacy for being physically active,(f) perceived support for eating healthfully, and (g) perceived support for being physically active will be included. Outcome measures include intentions to toward eating healthfully, intentions towards physical activity, current eating activity and current physical activity as well as health status measures of BMI z-scores and measured blood pressure. Data will be analyzed using multi-level linear models. The models will be augmented with an additional independent variable of the number of sessions attended by the child, entered as a moderating co-variate. Each outcome measure (intentions, current behaviors and health status) will be modeled separately. Independent variables will include the treatment group (between subjects), time period (within subjects), mentor/adult group leader ID (random effect), child ID and (random effect nested within mentor/adult group leader). Grade, gender, school and baseline z-scores will be included to account for the stratification in the experimental design, except in the model for BMI. Interaction effects will be tested. Results will indicate if the teen mentoring approach is more effective in promoting healthy eating and physical activity as the traditional adult group leader approach. If successful, teen mentoring may be a more effective way to promote healthy behaviors among children in poor rural areas with very limited resources for formal adult- mediated health promotion programs. PUBLIC HEALTH RELEVANCE: Obesity in children is a national epidemic, but certain subgroups are at higher risk. This study targets one high risk subgroup, children living in a rural, low-income Appalachian area. The study will test an innovative mentoring model that pairs school aged children with trained high school teen mentors to promote the adoption of healthier eating patterns and regular physical activity. The mentoring delivery method is compared to the usual condition of using adult group leaders in a classroom setting. If successful, teen mentoring may be a more effective way to promote healthy behaviors among children particularly in poor rural areas with very limited resources for formal health promotion programs.